Harmonic Mitigation in Large Industrial Plants

[Rodmcm]

I have a client that has large block loads on an industrial plant that each have high VFD driven harmonics. He has implemtented mitigation ( active harmonic units) for one block and now is thinking about the second block to be commissioned next year. Using phase shifting power transformers (relative to first block) is a cheap solution but if he loses load on one block then the mitigation effect on the other is reduced. A thought was to sense the harmonics at the incomer (point of common coupling to the two load blocks) and use this to remotely control the active harmonic units in each block. The present technology does not seem to allow for this. Does anyone have any experience in this area?

Also I am looking for a 'simple' software package that can simulate the additive effects of harmonics when phase shifted. Does any thing exist? Matlab?

 

[Rodmcm]

Also needed to say that the blocks are not the same load and hence some active filters will be left to account for this misbalance. These are the ones we could use for mitigation at low load but would be spread over two blocks.

 

[PHovnanian]

Active harmonic filters and phase shifting are two different approaches to harmonic mitigation, each with its own issues.

Active filters adjust their harmonic injection based upon system loads to cancel harmonic demands. So I would expect that they provide their own sensing and control algorithms and additional equipment is not needed.

Phase shifting uses the harmonic demands of one load block to cancel those of another. If the combination of connected load blocks varies, then the total harmonic loads will as well. Power up the 'wrong' combination of loads and a phase shifting scheme may provide no improvement for that load block. However, will all loads powered, phase shifting will provide cancellation to hold the maximum harmonic demand below the uncompensated level.

Attempting to control a phase shifting mitigation scheme will require switching transformers in or out based on some control algorithm. And that will require full power load switching, which won't be cheap and could disrupt load operation (break transfers, etc.). So phase shifting schemes are usually fixed at design time.

Combining the two schemes could work. Phase shifting to hold the maximum harmonic demand down to that of one load subset's worst case plus active cancellation sized to mitigate this (lower) level.

 

[Rodmcm]

Thank you. We were not considering switching power transformers only the harmonic filters. There has to be some in the system as the loads are not matched and we need a few of them to suck up the 'balance'. These would then become controllable if one block load decrease to assist with mitigation of the running block

 

[PHovnanian]

I would think that placing your active filter on the line side of your phase shift transformer feeds should be sufficient. And that active filter's sensing point would be at its point of connection to the system. No remote sensing required.

Looking at this the other way around, you want your phase shift transformers to be fed downstream of this filter, with no further filtering on their load side. The reason being: the phase shifting effect cancels its load harmonics with those of an adjacent load. So filtering harmonics further downstream of these phase shifts leaves you nothing to cancel with.

Furthermore, you don't want your active filter sensing to be remote from the filter itself where 'remote' means on the other side of a phase shift transformer. This could result in improper operation of the filter units (ask the mfg. about this) since they aren't reading the same harmonic VAs that they are injecting.

 

[ghostbuster7]

Also you may want to take a look at the lineator filter unit for drives from Mirus

 

[itsmoked]

Why not just put harmonic filters in front of each VFD then only spec VFDs with PFC for all future replacements?

Keith Cress
kcress -

http://www.flaminsystems.com

 

[ozmosis]

Rodmcm
I would suggest asking the supplier of the drives to assist with simulation sw. Most companies should provide something and generally will provide some tools free.
As a guide (only), our company provides this tool FOC for running simulations:

http://www.danfoss.com/BusinessAreas/DrivesSolutions/Software+MCT31/

It uses a number of assumptions so cannot be relied upon as guaranteed in any way, but should give you an idea.
Note: this is only for low voltage drives (<inc 690V) and all drives include an integral DC link choke that already brings THID down to approx 40%.